How are clustering and classification different in data mining?

In data mining, the primary difference between clustering and classification lies in whether the target groups are known in advance.

Classification

Classification involves assigning objects to preexisting groups based on established statistical rules or models. It is a supervised approach where the categories are defined before the analysis begins.1

Clustering

Clustering, often referred to as descriptive modeling, is an inductive, exploratory technique used to discover hidden structures or patterns within data.21
* Group Discovery: Unlike classification, the proper groups are not known in advance; the algorithm determines the groupings by analyzing the data itself.21
* Similarity: It organizes objects into groups where similarity is maximized within a group and minimized between different groups.1
* Hypothesis Generation: It is considered a hypothesis-generating technique rather than a hypothesis-testing one, as it uncovers structures without explaining the underlying reasons for their existence.1

Common clustering methods include hierarchical algorithms and nonhierarchical techniques like the k-means algorithm, which seeks to minimize variability within clusters while maximizing it between them.1